The molecular level modeling of MHC II-peptide systems to obtain a solid theory of the origins of autoimmune systems, particularly rheumatoid arthritis (RA) is proposed. The first modeling efforts will be directed toward a systematic search of the structure of a "kinetic" isomer of an MHC-peptide system where a long and a short lived isomer have been observed (MHC II I-E/kPiggeon Cyt c peptide). Then, a hypothesis has been laid out in the proposal where a specific kinetic isomer of a collagen peptide in HLA-DR4 (DRB1 0401 allele) is not recognized as "self" by the immune system. The T-cells recognizing the "kinetic" isomer may not have been eliminated during the early maturation stage. Modeling of the proposed isomer will be done and may be eventually confirmed through experimental techniques at the McConnell lab. Specifically, it is proposed that water molecules solvating the two carboxylates in the P6 pocket of HLA-DR4 may interfere with the deep burial of the P6 Ala of the collagen peptide observed in the MHC-peptide crystal structure.